clarify-goals

Clarify a vague request into purpose, constraints, success criteria, unknowns, and next-step readiness. Activate before planning or implementation when the real objective is still blurry.

Safety Notice

This listing is from the official public ClawHub registry. Review SKILL.md and referenced scripts before running.

Copy this and send it to your AI assistant to learn

Install skill "clarify-goals" with this command: npx skills add malleus35/agora-clarify-goals

Clarify Goals

Purpose

Turn a vague request into an actionable brief.

This skill exists to prevent false starts. Do not proceed into planning, coding, or debate until the core decision is clear enough to state plainly.

Activate when

Use this skill when any of the following are true:

  • the request is broad, fuzzy, or overloaded
  • the user asks for a solution before the underlying goal is clear
  • multiple success criteria are implied but not distinguished
  • the team is jumping into execution with different problem frames
  • a debate is starting, but the actual decision is not yet visible

Do not use when

Skip this skill when:

  • the task already has a crisp decision question
  • constraints and success criteria are explicit
  • the user is asking for straightforward execution on a settled plan

Required questions

Always extract answers for these five fields:

  1. Purpose
  • What must be different as a result of this work?
  • What real problem is being solved?
  1. Constraints
  • What limits the space? Time, risk, dependencies, policy, tooling, social cost.
  1. Success criteria
  • What observable outcome would count as success?
  • What would Descartes later be able to verify?
  1. Unknowns
  • What materially important facts are missing?
  • Which assumptions are currently doing too much work?
  1. Decision posture
  • Is this mainly a clarification, decision, ideation, review, or governance problem?

Procedure

Step 1 — Restate the request plainly

Rewrite the request in one sentence without embellishment. If you cannot do this, the request is not yet clear.

Step 2 — Separate goal from solution

Identify where the user has stated a preferred solution before proving it is the right problem frame. Call this out explicitly.

Step 3 — Extract the five fields

Fill Purpose, Constraints, Success Criteria, Unknowns, Decision Posture.

Step 4 — Identify the blocking ambiguity

Name the one ambiguity that most threatens good judgment if ignored.

Step 5 — Recommend next workflow

Route to one of:

  • frame-the-decision
  • compare-options
  • doubt-list
  • assumption-audit
  • minority-report
  • court-review
  • dialectic overlay
  • skeptic overlay
  • genealogy overlay
  • court overlay

Output artifact

## Clarification Brief

### Request in Plain Language
- ...

### Purpose
- ...

### Constraints
- ...

### Success Criteria
- ...

### Unknowns
- ...

### Decision Posture
- Clarification / Decision / Ideation / Review / Governance

### Blocking Ambiguity
- ...

### Recommended Next Step
- Skill or overlay: ...
- Why: ...

Guardrail

Do not proceed into implementation or deep planning if any of these are still missing:

  • a concrete purpose
  • at least one success criterion
  • the main blocking ambiguity

If clarity is insufficient, stop and say so. A fast wrong start is worse than a slow correct frame.

Failure modes

Common failure modes:

  • treating a requested solution as the real goal
  • listing vague aspirations instead of measurable success criteria
  • flattening multiple decisions into one blob
  • hiding key unknowns because they are inconvenient
  • routing directly to dialectic when simple clarification would suffice

Escalation points

Escalate to the user when:

  • competing purposes cannot both be optimized
  • the key constraint is political or organizational rather than technical
  • success criteria conflict with each other
  • the user must choose which ambiguity to resolve first

Completion condition

This skill is complete only when the output makes the next workflow obvious. If the next step is still ambiguous, clarification is incomplete.

Source Transparency

This detail page is rendered from real SKILL.md content. Trust labels are metadata-based hints, not a safety guarantee.

Related Skills

Related by shared tags or category signals.

General

通义晓蜜 - 智能外呼

触发阿里云晓蜜外呼机器人任务,自动批量拨打电话。适用于批量外呼、客户回访、满意度调查、简历筛查约面试等场景。可从前置工具或节点获取外呼名单。

Registry SourceRecently Updated
General

Letterboxd Watchlist

Scrape a public Letterboxd user's watchlist into a CSV/JSONL list of titles and film URLs without logging in. Use when a user asks to export, scrape, or mirror a Letterboxd watchlist, or to build watch-next queues.

Registry SourceRecently Updated
General

Seedance Video Generation

Generate AI videos using ByteDance Seedance. Use when the user wants to: (1) generate videos from text prompts, (2) generate videos from images (first frame, first+last frame, reference images), or (3) query/manage video generation tasks. Supports Seedance 1.5 Pro (with audio), 1.0 Pro, 1.0 Pro Fast, and 1.0 Lite models.

Registry SourceRecently Updated
4.2K17jackycser
General

Universal Skills Manager

The master coordinator for AI skills. Discovers skills from multiple sources (SkillsMP.com, SkillHub, and ClawHub), manages installation, and synchronization...

Registry SourceRecently Updated